DESCRIPTION: The vast majority of cost and lost productivity in workers' compensation is due to work related --musculoskeletal injuries. Among injured workers with these conditions, a small proportion (5-10 percent) develops long-term disability and account for most (80-85 percent) of the cost and lost work. In the absence of an accurate method to identify workers at risk for long-term disability, secondary prevention efforts cannot be well targeted. This is a 5-year, population-based, prospective study among Washington State workers with back injuries and carpal tunnel syndrome (CTS). The principal aim is to develop an accurate predictive model of risk for long-term disability among 5 key risk dimensions: employment related factors; biomedical healthcare related factors; socio-demographic factors; administrative/legal factors; and psychosocial factors. To accomplish this aim, a baseline interview among approximately 3000 eligible workers will be conducted within 2-6 weeks of (workers compensation) claim allowance. A continuous measure of disability outcome (lost time compensation) will be determined from a computerized database at one year. Additional important outcomes will be determined by a follow-up interview at one year (functional status, work status) and from computerized records (work status, wage status). Multivariate survival analysis within and across risk dimensions will be used to develop the principal risk models, including adjustment for injury severity. The reliability of determining severity from medical records will be determined as well (Specific Aim 2). A third aim of this study is to develop a brief risk assessment instrument for both low back and CTS injuries which would be useful to physicians when first treating injured workers. The main focus of these instruments would be on those risk factors which may be modifiable and which may be amenable to early intervention to prevent disability. Statistical analysis for this aim will focus on sensitivity and specificity of specific combinations of risk factors. These risk assessment instruments will be pilot tested among physicians participating in an occupational health care quality improvement project. The fourth aim of this project is to determine reinjury rate and predictors of reinjury among the original low back injury cohort 2 years after the initial injury. The unique study environment in Washington state, including the ability to conduct population-based sampling, timely access to survey injured workers, linkage of medical, claims and work status databases, and support from the business and labor communities, will substantially contribute to the success of this study.